[hts-users:00952] Re: question about feature extraction

["Heiga ZEN (Byung Ha CHUN)" <zen@xxxxxxxxxxxxxxx>]

Hi,

Yufeng Hao wrote (2007/11/19 18:20):

> recently, I made some tries on HTS training under version 2.0.1 with my 
> cmu arctic like english voice.
>
> Configurations are list below:
> ./configure --with-tcl-search-path=/opt/ActiveTcl- 8.4/bin \
>                  
> --with-fest-search-path=/home/HMM/festival/festival/examples \
>                  --with-sptk-search-path=/usr/local/SPTK/bin \
>                  --with-hts-search-path=/usr/local/bin \
>     MGCORDER=12 MGCLSP=1 GAMMA=3 FREQWARP=0.42 LNGAIN=1
>
>
> while feature extracting process, some warning message displays, but I 
> can not find related explainings.
> The warning message list as below:
>
> Extracting MGC-LSP coefficients from raw/cmu_us_arctic_slt_a0289.raw
> mgcep : Error in theq() at 0th iteration!
> Extracting MGC-LSP coefficients from raw/cmu_us_arctic_slt_a0290.raw
> Extracting MGC-LSP coefficients from raw/cmu_us_arctic_slt_a0291.raw
> Extracting MGC-LSP coefficients from raw/cmu_us_arctic_slt_a0292.raw
> mgcep : Error in theq() at 0th iteration!
> Extracting MGC-LSP coefficients from raw/cmu_us_arctic_slt_a0293.raw
> Extracting MGC-LSP coefficients from raw/cmu_us_arctic_slt_a0294.raw
>
> Does anybody tell me what happened? Is there any harmful for following 
> training?

It means that determinant of Toeplitz+Hankel matrix, which is a coefficient matrix of a linear system in MGC analysis, is too small (<1.0e-6).
You can modify this threshold via -f option as

mgcep -a 0.42 -g 3 -m 12 -f 0.01 ...

Please run it again with a smaller threshold.

Regards,

--
------------------------------------------------
Heiga ZEN     (in Japanese pronunciation)
Byung Ha CHUN (in Korean pronunciation)

Department of Computer Science and Engineering
Nagoya Institute of Technology
Gokiso-cho, Showa-ku, Nagoya 466-8555 Japan

http://www.sp.nitech.ac.jp/~zen
------------------------------------------------